© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Jacobian LensNEW
    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsinterp-engineNEWAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Under Peer Review · Sparse Autoencoders for Pythia-70M-Deduped
    5. Pythia-70M-Deduped
    7. Residuals
    9. 2-RES-SM
    11. 26525
    Prev
    Next
    INDEX
    Explanations

    semicolons and punctuation marks

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Configuration
    ctigges/pythia-70m-deduped__res-sm_processed/2-res-sm
    Prompts (Dashboard)
    32,768 prompts, 128 tokens each
    Dataset (Dashboard)
    monology/pile-uncopyrighted
    Features
    32,768
    Data Type
    torch.float32
    Hook Name
    blocks.2.hook_resid_post
    Hook Layer
    2
    Architecture
    standard
    Context Size
    128
    Dataset
    EleutherAI/the_pile_deduplicated
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    34
    +0.13
    0.8%
    222
    +0.13
    0.8%
    263
    +0.13
    0.7%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    222
    +0.13
    0.02
    192
    +0.13
    0.01
    315
    +0.13
    0.01
    Negative Logits
    »¿
    -1.77
    terior
    -1.56
     respectively
    -1.54
     respective
    -1.54
    ptive
    -1.52
    iration
    -1.43
    endo
    -1.43
    plastic
    -1.40
    astom
    -1.37
    ytes
    -1.37
    POSITIVE LOGITS
     DISCLAIM
    1.58
    ÂĹ
    1.44
    ]{.
    1.40
     avail
    1.33
     -*-
    1.30
     ÂĹ
    1.29
    llll
    1.27
     promptly
    1.27
     worms
    1.26
     excuse
    1.26
    Activations Density 0.084%

    No Known Activations